Unbiased curriculum learning enhanced global-local graph neural network for protein thermodynamic stability prediction.

MOTIVATION: Proteins play crucial roles in biological processes, with their functions being closely tied to thermodynamic stability. However, measuring stability changes upon point mutations of amino acid residues using physical methods can be time-consuming. In recent years, several computational methods for protein thermodynamic stability prediction (PTSP) based on deep learning have emerged. Nevertheless, these approaches either overlook the natural topology of protein structures or neglect the inherent noisy samples resulting from theoretical calculation or experimental errors. RESULTS: We propose a novel Global-Local Graph Neural Network powered by Unbiased Curriculum Learning for the PTSP task. Our method first builds a Siamese graph neural network to extract protein features before and after mutation. Since the graph's topological changes stem from local node mutations, we design a local feature transformation module to make the model focus on the mutated site. To address model bias caused by noisy samples, which represent unavoidable errors from physical experiments, we introduce an unbiased curriculum learning method. This approach effectively identifies and re-weights noisy samples during the training process. Extensive experiments demonstrate that our proposed method outperforms advanced protein stability prediction methods, and surpasses state-of-the-art learning methods for regression prediction tasks. AVAILABILITY AND IMPLEMENTATION: All code and data is available at https://github.com/haifangong/UCL-GLGNN.

Thyroid region prior guided attention for ultrasound segmentation of thyroid nodules.

Ultrasound segmentation of thyroid nodules is a challenging task, which plays an vital role in the diagnosis of thyroid cancer. However, the following two factors limit the development of automatic thyroid nodule segmentation algorithms: (1) existing automatic nodule segmentation algorithms that directly apply semantic segmentation techniques can easily mistake non-thyroid areas as nodules, because of the lack of the thyroid gland region perception, the large number of similar areas in the ultrasonic images, and the inherently low contrast images; (2) the currently available dataset (i.e., DDTI) is small and collected from a single center, which violates the fact that thyroid ultrasound images are acquired from various devices in real-world situations. To overcome the lack of thyroid gland region prior knowledge, we design a thyroid region prior guided feature enhancement network (TRFE+) for accurate thyroid nodule segmentation. Specifically, (1) a novel multi-task learning framework that simultaneously learns the nodule size, gland position, and the nodule position is designed; (2) an adaptive gland region feature enhancement module is proposed to make full use of the thyroid gland prior knowledge; (3) a normalization approach with respect to the channel dimension is applied to alleviate the domain gap during the training process. To facilitate the development of thyroid nodule segmentation, we have contributed TN3K: an open-access dataset containing 3493 thyroid nodule images with high-quality nodule masks labeling from various devices and views. We perform a thorough evaluation based on the TN3K test set and DDTI to demonstrate the effectiveness of the proposed method. Code and data are available at https://github.com/haifangong/TRFE-Net-for-thyroid-nodule-segmentation.

Development and validation of a novel necroptosis-related score to improve the outcomes of clear cell renal cell carcinoma.

Necroptosis has been indicated as a key regulator of tumor progression. However, the prognostic regulatory role of necroptosis in clear cell renal cell carcinoma (ccRCC) needs to be further investigated. In this study, necroptosis-related subtypes were identified by mining the public cohort (n = 530) obtained from The Cancer Genome Atlas. By applying Principal Component Analysis (PCA), the necroptosis-related scores (N-Score) were developed to assess the prognosis procession of ccRCC. The results were further validated by an external clinical cohort (n = 116) obtained from the First Affiliated Hospital of Wenzhou Medical University. It has been found that N-Score could precisely distinguish the prognostic outcomes of patients as an independent risk factor (Hazard ratio = 4.990, 95% confidence interval (CI) = 2.007-12.403, p < 0.001). In addition, changes in N-Score were associated with differences in tumor mutational burden as well as immune infiltration characterization. Moreover, higher N-Scores were also correlated significantly molecular drug sensitivity and stronger immune checkpoint activity. Notably, the prognosis of ccRCC could be effectively guided by combining the N-Scores and external clinical indicators. In conclusion, N-Scores could be served as a robust and effective biomarker to improve the prognosis outcomes and targeted therapy of ccRCC.

VQAMix: Conditional Triplet Mixup for Medical Visual Question Answering.

Medical visual question answering (VQA) aims to correctly answer a clinical question related to a given medical image. Nevertheless, owing to the expensive manual annotations of medical data, the lack of labeled data limits the development of medical VQA. In this paper, we propose a simple yet effective data augmentation method, VQAMix, to mitigate the data limitation problem. Specifically, VQAMix generates more labeled training samples by linearly combining a pair of VQA samples, which can be easily embedded into any visual-language model to boost performance. However, mixing two VQA samples would construct new connections between images and questions from different samples, which will cause the answers for those new fabricated image-question pairs to be missing or meaningless. To solve the missing answer problem, we first develop the Learning with Missing Labels (LML) strategy, which roughly excludes the missing answers. To alleviate the meaningless answer issue, we design the Learning with Conditional-mixed Labels (LCL) strategy, which further utilizes language-type prior to forcing the mixed pairs to have reasonable answers that belong to the same category. Experimental results on the VQA-RAD and PathVQA benchmarks show that our proposed method significantly improves the performance of the baseline by about 7% and 5% on the averaging result of two backbones, respectively. More importantly, VQAMix could improve confidence calibration and model interpretability, which is significant for medical VQA models in practical applications. All code and models are available at https://github.com/haifangong/VQAMix.

The Diagnostic Efficacy of the American College of Radiology (ACR) Thyroid Imaging Report and Data System (TI-RADS) and the American Thyroid Association (ATA) Risk Stratification Systems for Thyroid Nodules.

BACKGROUND: This study is aimed at evaluating the diagnostic efficacy of ultrasound-based risk stratification for thyroid nodules in the American College of Radiology (ACR) Thyroid Imaging Reporting and Data System (TI-RADS) and the American Thyroid Association (ATA) risk stratification systems. METHODS: 286 patients with thyroid cancer were included in the tumor group, with 259 nontumor cases included in the nontumor group. The ACR TI-RADS and ATA risk stratification systems assessed all thyroid nodules for malignant risks. The diagnostic effect of ACR and ATA risk stratification system for thyroid nodules was evaluated by receiver operating characteristic (ROC) analysis using postoperative pathological diagnosis as the gold standard. RESULTS: The distributions and mean scores of ACR and ATA rating risk stratification were significantly different between the tumor and nontumor groups. The lesion diameter > 1 cm subgroup had higher malignant ultrasound feature rates detected and ACR and ATA scores. A significant difference was not found in the ACR and ATA scores between patients with or without Hashimoto's disease. The area under the receiver operating curve (AUC) for the ACR TI-RADS and the ATA systems was 0.891 and 0.896, respectively. The ACR had better specificity (0.90) while the ATA system had higher sensitivity (0.92), with both scenarios having almost the same overall diagnostic accuracy (0.84). CONCLUSION: Both the ACR TI-RADS and the ATA risk stratification systems provide a clinically feasible thyroid malignant risk classification, with high thyroid nodule malignant risk diagnostic efficacy.

Depthwise Nonlocal Module for Fast Salient Object Detection Using a Single Thread.

Recently, deep convolutional neural networks have achieved significant success in salient object detection. However, existing state-of-the-art methods require high-end GPUs to achieve real-time performance, which makes it hard to adapt to low cost or portable devices. Although generic network architectures have been proposed to speed up inference on mobile devices, they are tailored to the task of image classification or semantic segmentation, and struggle to capture intrachannel and interchannel correlations that are essential for contrast modeling in salient object detection. Motivated by the above observations, we design a new deep-learning algorithm for fast salient object detection. The proposed algorithm for the first time achieves competitive accuracy and high inference efficiency simultaneously with a single CPU thread. Specifically, we propose a novel depthwise nonlocal module (DNL), which implicitly models contrast via harvesting intrachannel and interchannel correlations in a self-attention manner. In addition, we introduce a depthwise nonlocal network architecture that incorporates both DNLs module and inverted residual blocks. The experimental results show that our proposed network attains very competitive accuracy on a wide range of salient object detection datasets while achieving state-of-the-art efficiency among all existing deep-learning-based algorithms.